Readings Flashcards
What are the conditions for a variable to be considered a covariate?
It must be (1) correlated with the dependent variable and (2) not systematically related to the independent variable.
Why should covariates be measured before exposing participants to experimental conditions?
To prevent the manipulation from influencing the covariate measurement.
What statistical tests can incorporate covariates?
ANCOVA and MANCOVA.
Why is realism important in experimental research?
It increases external validity and ensures findings generalize to real-world settings.
What are the two dimensions of realism in experiments?
Experimental realism (how engaging the task feels) and mundane realism (how similar the setting is to real life).
What are ways to enhance realism in an experiment?
Using real-world stimuli, measuring actual behaviors, and designing immersive scenarios.
What type of dependent variable is most realistic in consumer behavior research?
Actual consumer behavior rather than self-reported intentions.
What are the advantages of a between-subjects design?
Avoids carryover effects, demand effects are lower, and simpler execution.
What are the disadvantages of a between-subjects design?
Requires more participants, increased variance due to individual differences, and lower statistical power.
What are the advantages of a within-subjects design?
Requires fewer participants, increases statistical power, and allows for direct comparisons.
What are the disadvantages of a within-subjects design?
Prone to carryover effects, demand effects, and learning effects.
What is statistical power?
The probability of correctly detecting a true effect when it exists.
What factors increase statistical power?
Larger sample sizes, stronger manipulations, and reducing noise (e.g., using covariates).
How does effect size impact power?
Larger effect sizes require smaller sample sizes to achieve the same level of power.
What reduces power in an experiment?
High measurement error, small effect sizes, unmotivated participants, and satisficing responses.
How can instructional manipulation checks (IMCs) improve power?
By identifying inattentive participants and reducing noise in the data.
What is satisficing in experimental research?
When participants provide minimally acceptable answers instead of thoughtful responses.
How does satisficing impact data quality?
It increases noise, reduces reliability, and weakens experimental effects.
What methods help detect satisficing?
Attention checks, response time tracking, and consistency tests across responses.
What are the trade-offs of using attention checks in experiments?
They improve data quality but may reduce sample size and increase participant dropout rates.
Why are manipulation checks important?
They confirm whether the experimental manipulation had the intended effect.
When should manipulation checks be conducted?
After the dependent variable but before demographics to avoid biasing responses.
What statistical tests are used for manipulation checks?
ANOVA for continuous measures and chi-square tests for categorical measures.
What is a confounding check?
A test to ensure that the manipulation only influenced the intended construct.
What is the trade-off between internal and external validity in experiments?
High internal validity ensures causal inference, but it may come at the cost of generalizability.
How can researchers balance internal and external validity?
By using a mix of lab and field studies, using real-world stimuli, and ensuring manipulations are ecologically valid.
What is orthogonality in experimental design?
The effects of different factors or treatments are independent and can be estimated separately, ensuring that the effect of one factor doesn’t influence the estimation of another
Why is orthogonality important in experiments?
It prevents confounding effects and allows for clear interpretation of independent variable effects.
What are the 4 goals of experimental consumer research according to Morales et al. (2017)?
- Identify causal relationships, 2. Test theoretical predictions, 3. Examine process mechanisms, 4. Extend findings to new contexts.
What are the 3 types of experiments outlined by Morales et al. (2017)?
- Theory-testing experiments, 2. Process-revealing experiments, 3. Managerially relevant experiments.
How do theory-testing experiments differ from process-revealing experiments?
Theory-testing experiments establish causality, while process-revealing experiments examine the underlying mechanisms of an effect.
What is the goal of managerially relevant experiments?
To provide insights that directly inform business decisions and strategy.
